Horangi 4 - Ko-HalluLens (WikiQA): leaderboard
Metric: Correct answer rate (%). Source: horangi.ai. 105 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.6 Sol (Max) | 64 |
| 2 | Gemini 3.1 Pro (Preview) (High) | 64 |
| 3 | GPT-5.5 (xHigh) | 63 |
| 4 | Gemini 3 Pro (Preview) (High) | 62 |
| 5 | Grok 4.5 | 60 |
| 6 | Gemini 3.5 Flash | 59 |
| 7 | Gemini 3 Flash (Preview) | 58 |
| 8 | Claude Fable 5 (High) | 55 |
| 9 | GPT-5.4 (xHigh) | 54 |
| 10 | GPT-5.6 Luna (Max) | 52 |
| 11 | GPT-5 | 51 |
| 12 | Gemini 3.1 Flash Lite (Preview) | 50 |
| 13 | GPT-5.1 | 48 |
| 14 | GPT-5.6 Terra (Max) | 48 |
| 15 | GPT-4.1 | 45 |
Interactive version: theaggregate.ai/benchmark?slug=horangi-4-ko-hallulens-wikiqa · How It Works · Data refreshed daily, snapshot 2026-09-19.